February 2, 1922] 



NATURE 



141 



\ 



Before entering into further discussions of the 

 facts, it will be convenient to mention a few of the 

 generally accepted views as to the constitution of 

 the terrestrial atmosphere and its ionisation by light. 



By the use of hydrogen-filled sounding balloons 

 carrying self-recording meteorographs, it has been 

 possible to explore the atmosphere up to a height of 

 about twenty miles. One of the results is to show 

 that our atmosphere may roughly be divided into 

 two regions. In the lower layer, called the tropo- 

 sfhere, the atmospheric gases are kept well mixed 

 up by winds and convection. This layer extends to 

 a height of six or seven miles or so, and in it the 

 temperature falls regularly with increasing height at 

 the rate of about 6° C. per kilometre of ascent until 

 a temperature of about —55° C. is reached. Above 

 this is a zone called the stratosphere, of unknown 

 thickness, in which the temperature remains con- 

 stant. Above a height of about seven miles water 

 vapour is absent, and at higher levels convection 

 ceases to operate and the atmospheric gases arrange 

 themselves in order of densitv. The outer and highest 

 levels above a height of sixty miles (=100 km.) 

 iare chieflv composed of helium and hydrogen with 

 possibly some small admixture of the rarer atmo- 

 spheric gases neon and krypton. 



The volume composition of the atmosphere at the 

 earth's surface is as follows : — 



Nitrogen ... 78-05 per cent. 



Oxygen ... 2i"oo ,, 



Argon ... 0-93 „ 



Carbon dioxide o 03 ,, 



Hydrogen ... i to 10 vols, in a million of air. 



Neon... ... 10 „ „ 



Helium ... i to 2 „ „ 



Krypton ... i vol. „ „ 



Xenon ... 0-05 „ „ ,, 



Oxygen is almost entirely absent at a height of 

 ^■100 km., but nitrogen is still present in a 

 rarefied form. The presence of hydrogen and helium 

 at these high levels has been indicated by an observa- 

 tion of Pickering on the spectrum of a meteoric 

 stone entering the earth's atmosphere, which showed 

 the hydrogen and helium lines. 



Next, as regards the action of light on the gases 

 of the atmosphere. Light waves of high refrangi- 

 bility impinging on nearly all substances, especially 

 those containing electropositive atoms, liberate from 

 them electrons. The atom is now considered to be 

 a collocation of negative electrons arranged in con- 

 centric shells, possibly in orbital motion, round a 

 central positively charged nucleus in which the gravi- 

 tative mass of the atom chiefly resides. Light of 

 short wave-length causes one or more of these nega- 

 tive electrons to be detached and projected with a 

 high velocity. The more electronegative an atom is 

 the higher must be the frequency of the light to 

 affect it. The electrons so detached are called photo- 

 electrons and the action photo-electric. 



In the case of sodium or potassium, which are 

 highly electropositive metals, photo-electrons are 

 emitted under the action of visible light, about the 

 middle of the spectrum, but for less electropositive 

 metals — e.g. zinc and magnesium — the action takes 

 NO. 2727, VOL. 109] 



place only with ultra-violet light. Hence it follows 

 that a plate of zinc illuminated by light from an 

 electric arc or by the spark between aluminium balls 

 loses a negative charge readily, and if insulated, 

 becomes positively electrified owing to the loss of 

 negative photo-electrons. The velocity with which 

 these photo-electrons are projected is considerable, 

 and may be 500—1000 km. per sec. 



For most metals the ionising potential is about 

 two to four volts, hence the maximum wave-length 

 of ionising light is just beyond {he violet end of 

 the visible spectrum. But for atmospheric gases, 

 when pure and free from dust or moisture, the ion- 

 ising potential is much higher, being approximately 

 as follows : — 



It follows from this that the atmospheric gases 

 cannot be ionised by light of longer wave-length 

 than 1350 A.U, Rays of this short wave-length 

 are not transmitted by quartz but only by certain 

 samples of fluorite, and are absorbed by a very 

 small thickness of air. No sunlight of shorter wave- 

 length than about 2950 A.U. reaches the earth's 

 surface, as shown long ago by Huggins and Cornu. 



Hence the conclusion is forced on us that pure 

 dust-free atmospheric gases cannot be ionised at the 

 lower levels of the atmosphere by the direct action 

 of sunlight, but at the higher levels above 60 to 

 100 km. doubtless there is direct ionisation. 



Nevertheless, ionisation does take place in the 

 lower atmospheric levels, as shown by the small 

 finite electric conductivity possessed by the air, 

 which proves that there are negative ions, either free 

 electrons, or electrons attached to neutral atoms, and 

 also positive ions present in the air, even over wide 

 oceans. Thus, Boltzmarm found in tests made in 

 mid-Atlantic 1150 positive and 800 negative ions per 

 c.c. of air. A. S. Eve found 600 to 1400 positive 

 and 500 to 1000 negative ions per c.c, the posi- 

 tive being slightly in excess. 



This ionisation may be produced either by photo- 

 electric action on dust or ice particles in the air, 

 by radio-active matter in the soil, by photo-electric 

 action upon complex gaseous molecules in the air, 

 or generated by the light and called condensation 

 nuclei. Such agencies, however, cannot account for 

 the far larger and permanent ionisation necessary to 

 give the required electric conductivity in the higher 

 atmosphere if it is to act as a guide to long electro- 

 magnetic waves. 



A consideration of the terrestrial radio-telegraph 

 problem shows that if there is any conductive layer 

 in the upper atmosphere which can act as a guide 

 to long electromagnetic waves round the earth, it 

 must possess the following properties : — ^ 



(i) It must be permanently ionised, which means 

 that its ionisation must not vanish in the night-time 

 since, so far as we know, its guiding powers are 

 not suspended on the shadow side of the earth. 

 This seems to imply that the ionisation must be 

 predominantly nf one sign or that the plus and 



